Various techniques have been developed for the targeted and untargeted analysis of complex mixtures using tandem mass spectrometry (MS).
The traditional approach for untargeted analysis (that is, analysis without prior knowledge) of an analyte is to carry out a data dependent selection of a suitable precursor ion of a particular mass to charge ratio (m/z). For example, the, or one of the, more intense peaks in the mass spectrum, which has not yet been analysed, can be selected. That suitable precursor can then be fragmented and the fragments detected in an MS/MS analysis technique.
Selection/isolation of the suitable precursor ion is typically achieved by a quadrupole mass filter or linear trap analyzer. Fragmentation of the selected precursor may be achieved, typically, through collision of the precursor ion with gas or ion-ion or ion-molecule reactions. The detection of the resulting fragments may be achieved through a scanning quadrupole filter or, in preference, by using an all-ion analyzer such as a time of flight (TOF), Orbitrap™ or Fourier Transform Ion Cyclotron Resonance (FTICR) analyzer.
A drawback of the above arrangement is that only a restricted number of available precursors will generate a corresponding MS/MS spectrum, as a result of limitations on transmission and the complexity of mixtures. In consequence, the depth of analysis of complex mixtures such as are found in proteomics, environmental, food, drug metabolism and other applications is severely curtailed.
An alternative to this traditional approach employs MS/MS but splits the ion beam from the ion source into packets according to their mass to charge ratio. A particular packet or packets is/are fragmented without loss of others of the packets, or alternatively, in parallel with other of the packets. This splitting into packets may be performed using a scanning device which stores ions of a broad mass range, such as a 3D ion trap as is disclosed, for example, in WO-A-03/103,010, or a linear trap with radial ejection as is disclosed in, for example, U.S. Pat. No. 7,157,698. Alternatively, packet splitting may be achieved using pulsed ion mobility spectrometry, and some suitable apparatuses and techniques are described in WO-A-00/70335 and US-A-2003/0,213,900 respectively. Still further alternatives involve slowed down linear mass spectrometers, see for example WO-A-2004/085,992, or multi reflection time of flight mass spectrometers as in WO-A-2004/008,481.
In all of the above cases, the first stage of mass analysis is followed by fast fragmentation, for example in a collision cell (preferably with an axial gradient), or using a pulsed laser. The fragments are then analysed, again in preference using another TOF mass spectrometer on a much faster timescale than the scanning duration (the fast analysis times are referred to in the art as “nested times”). The overall performance is, however, compromised because only a very limited time is allocated to each scan (typically, no more than 10-20 microseconds).
These approaches of so called “two dimensional MS” apparently provide improved throughput without comprising sensitivity. In this respect they are superior to a variant of traditional MS/MS, expanded to a multi channel configuration in which a number of parallel mass analyzers (typically ion traps) are used to select one precursor each, and then its fragments are scanned out to an individual associated detector (eg the ion trap array of U.S. Pat. No. 5,206,506 or multiple traps of US-A-2003/089,846).
Even so, all 2D-MS techniques currently representing the state of the art suffer from relatively low resolution of precursor selection (typically, no better than one to several atomic mass units, a.m.u.). They also tend to suffer from relatively low resolving power of fragment analysis—typically no better than a few hundred to a few thousand (and thus provide poor mass accuracy). Furthermore, the known 2D-MS techniques are each based on the use of trapping devices to provide a high duty cycle. Such devices have an overall cycle time which is defined by the cycle time of the slowest analyzer in the system. Modern ion sources produce ion current up to 100 s of pA, that is, in excess of 109 elementary charges per second. Thus, if the full cycle of scanning through the entire mass range of interest is 5 milliseconds, then such trapping devices need to be able to accumulate up to 5 million elementary charges yet still allow efficient precursor selection. These difficulties have precluded such approaches from entering main stream, practical mass spectrometry.
As a compromise, therefore, an alternative method has been developed on the basis of the time of flight (TOF) analyzer, and is available on the market under the name MSe. In this approach, precursor ions are caused to pass through a fragmentation or reaction device alternately at higher and lower energy, resulting in the formation of product ions in the former case (see, for example, U.S. Pat. Nos. 6,586,727 and 6,982,414). This can readily be accomplished using a Q-TOF type instrument, by operating the quadrupole mass filter in the RF-only mode such as the simultaneously transmit approximately a decade in mass into the gas collision cell with higher collision energy, sufficient to induce fragmentation. The technique is set out in for example Bateman et al., J Am Soc Mass Spectrom. 2002, 13, pages 792-803. The orthogonal time of flight mass spectrometer records the mass spectrum of the resulting mixture of precursor and fragment ions. It is not necessary to remove the gas from the collision cell. Hence, by alternating the collision energy (typically, from less than 10V to between 30 and 70V), it is possible to alternate between recording the spectrum exhibiting mainly precursor ions, and the spectrum exhibiting the mixture of precursor ions and their fragment ions.
In an alternative method to alternating the collision energy, ions may be directed into the fragmentation cell at an appropriate energy such that significant fragmentation occurs and from there to analysis. As a further alternative, ions may be allowed to enter the analyzer directly along a different path where significant fragmentation does not occur. Such a method is described in U.S. Pat. No. 7,759,638.
In the first mode, wherein relatively low collision energy is employed, no—or substantially no—fragmentation of ions takes place so that precursor ions will be relatively more intense in the resultant mass spectrum. In the second mode, wherein a relatively higher collision energy is employed, most or indeed all of the precursor ions are fragmented so that the fragment ions are relatively more intense in the resultant mass spectrum in this second mode. Hence, by suitable adjustment of the collision energy in the two operating modes, precursor and product ions may be readily distinguished. The method may be further enhanced by utilising the chromatographic separation of analytes which introduces a temporal dimension as well. That is, the method may utilise the dependence of ion current on retention time. From this, it is possible to group elution profiles of various fragment ions, with those of precursors, and thus in turn it is possible to separate one family of precursor ions, with its fragments, from another family of precursor ions. Furthermore, the use of high resolution/accurate mass analyzers makes such a grouping much more reliable.
Nevertheless, the MSe approach proposed by Bateman and others suffers from a number of limitations. Firstly, the extremely large number of precursors, and the range of their concentrations, in modern mass spectrometric analysis, limits the applicability of this method to the most intense peaks only: spectra become very crowded at lower intensities upon fragmentation. Secondly, there is no way to distinguish co-eluting peaks, which results in an increased number of false identifications, for complex mixtures. Thirdly, in consequence of the above, the method does not work for infusion, when no chromatographic peaks are formed. Fourthly, the high-energy fragmentation spectra typically exhibit many more peaks than the low-energy (non-fragmentation) spectra and can suffer from overcrowding of the spectra. The latter is especially pronounced when analyzing a single class of analytes such as peptides, which are all built from common aminoacids.
WO-A-2010/120496 describes an arrangement in which a multiple fill Higher Collision Energy Dissociation (HCD) cell functionality, or a C-trap cell functionality of an accurate-mass mass analyzer system is employed to avoid performing a separate full scan MS event. Instead a scan event is substituted which detects all ions originating from high and low collision energy fills simultaneously. This simultaneous analysis technique allows execution of all ion MS2 experiments significantly faster than when discrete spectra are acquired at specified collision energy. However, this method may still yield spectra that are more crowded that is desirable.